The health-related, social, and financial costs of drug abuse are well-known;these costs are only multiplied when perinatal drug use and associated risks (such as HIV) also place an infant at risk, both prenatally and postnatally. A range of relevant treatment programs exist, but such programs are not able to either reach all women who are in need of treatment, or to engage many of those they do identify and intend to treat. Even brief motivational interventions in primary care - despite tremendous promise - are limited by the amount of time, training, expertise, and interest they require from primary care providers. In contrast, computer-based brief interventions could allow access to high proportions of women using drugs during the perinatal period. Computers could also do so at very low cost, and with perfect replication in the community. A computer-based brief motivational intervention developed under a Stage I R21 has shown early promise: at 3-month follow-up in a preliminary clinical trial, postpartum women randomly assigned to a 20-minute computer-based brief intervention were 3 times less likely to be using either cocaine, other stimulants, or opiates. In this Stage II application, this research will be expanded in two ways. First, the intervention will be modified based on data from the preliminary clinical trial, and tested in a larger sample of drug-using postpartum women. Second, the intervention will be extended to also address problem alcohol use and prenatal smoking. Specifically, this research will: (a) revise and upgrade software based on data from the initial clinical trial;(b) recruit 500 post-partum women (200 screening positive for illicit drug use, 150 screening positive for problem alcohol abuse, and 150 screening positive for smoking) from an urban obstetric hospital;and (c) randomly assign women into intervention and control conditions, with blinded follow-up assessments at 3- and 6-months. In addition to testing the effect of the intervention on substance use outcomes (as measured by self-report and toxicological analysis) and related outcomes (such as HIV risk), data analyses will also examine theory-driven mediators and moderators of any observed association between the brief intervention and outcomes. This research could enhance public health by making available an effective, practical, low-cost, and high-reach intervention for all forms of perinatal substance abuse.